[unreadable] We are proposing a translational investigation, including clinical as well as basic scientists, to test the hypothesis that specific Pseudomonas virulence products determine whether lung infection occurs. The results from the proposed experiments will be used to develop diagnostic tools that can identify virulent Pseudomonas in patients. The results will also identify critical bacterial gene products involved in lung injury. The present application carefully dissects associations between bacterial virulence and lung injury so that future mechanistic investigations in patients can be proposed. Without the data from the proposed investigations, we do not have the information to proceed. We already have Pseudomonas strains obtained from a cohort of intubated, mechanically ventilated patients from another grant. Our first aim is to perform a genomic analysis of Pseudomonas strains obtained from 2 distinct patient populations; intubated, ventilated patients with signs of lung infections that have disseminated and intubated, ventilated patients who have no sign of disease. Using microarray analysis, we will determine the significantly different genes between the Pseudomonas strains found in these two patient groups and produce consensus primers to these genes. We will then utilize these consensus primers in a prospective investigation of 300 intubated, ventilated patients to confirm the prevalence of the genetic differences in colonized vs infected intubated, ventilated patients. The second aim is to compare the in vivo virulence gene expression of Pseudomonas strains obtained from colonized patients to the expression profile of the strains obtained from patients who have Pseudomonas lung infections Analysis of bacterial gene expression in vivo will be done using real-time PCR. The results of all these experiments will allow us to prove (or disprove) that there are different genes or different patterns of toxin gene expression between the Pseudomonas strains infecting patients and those colonizing patients. In the final aim, we will evaluate the interaction of type III toxin genes in vivo By utilizing an isogenic bacterial strain (PA103) that has each of the toxin genes added alone or in combination, we will be able to define the effects of each of the type III toxins as well as other genes we find of significance in the first two aims. [unreadable] [unreadable]